ABSTRACT
PURPOSE: Many investigators have reported the financial conflicts of interest (COI), which could result in potential bias in the reporting of outcomes for patients undergoing total disc replacement (TDR) rather than anterior cervical discectomy and fusion (ACDF). This bias may be subconsciously introduced by the investigator in a non-blinded radiographic review. The purpose of this study was to determine if bias was present when a group of spine specialists rated adjacent segment degeneration (ASD) following cervical TDR or ACDF. METHODS: Potential bias in the assessment of ASD was evaluated through the reviews of cervical radiographs (pre- and 6 years post-operative) from patients participating in the ProDisc-C FDA trial (ProDisc-C IDE #G030059). The index level was blinded on all radiographs during the first review, but unblinded in the second. Five reviewers (a radiologist, two non-TDR surgeons, and two TDR surgeons), two of whom had a COI with the ProDisc-C trial sponsor, assessed ASD on a three point scale: yes, no, or unable to assess. Intra- and inter-rater reliabilities between all raters were assessed by the Kappa statistic. RESULTS: The intra-rater reliability between reviews was substantial, indicating little to no bias in assessing ASD development/progression. The Kappa statistics were 0.580 and 0.644 for the TDR surgeons (p < 0.0001), 0.718 and 0.572 for the non-TDR surgeons (p < 0.0001), and 0.642 for the radiologist (p < 0.0001). Inter-rater reliability for the blinded review ranged from 0.316 to 0.607 (p < 0.0001) and from 0.221 to 0.644 (p < 0.0001) for the unblinded review. CONCLUSIONS: The knowledge of the surgical procedure performed did not bias the assessment of ASD.
